Quote:
Originally Posted by idoit
So I'm playing with init.d danger zone and you might already know Kindles are extremely fragile. As we don't have SSH access in the diags of DX(G), I thought of this. For scripts I want in /etc/init.d/, I just make a symlink to /mnt/us/ so that if a stupidity happens in my script, I still would have a chance to go to diag > mount usb > and comment the offending lines in the init.d script... Do you think, it is a good idea / safe?
I thought it's a clever idea, till we get Diag+SSH, eh?
For instance:
Code:
ln -s /mnt/base-us/applications/etc/init.d/myinitscript /etc/init.d/myinitscript
|
Perhaps a variation on that -
Quote:
|
/etc/init.d/rc.local is run very early in the initialization process and at every runlevel change.
|
Of course, that was from the mainstream init docs and the Kindles use the Busybox init - -
But if Busybox init still acts that way (about running an rc.local) - -
then linking it to: /mnt/base-us/esys/etc/init.d/rc.local should put it into user storage.
- - - -
Note:
What you read here:
https://bitbucket.org/twobob/kual-sy...92e6?at=master
Had a directory tree change:
From: /mnt/base-us/extensions/system
To: /mnt/base-us/esys
(Paths where getting too long for the ELF headers with the old naming plan.)
I should have (did have) owner/admin status on that repo - but am having trouble getting back in as admin.
That problem is on my todo list of things I need to fix or clone that repo somewhere else.